Session: Beyond the Beaver Tail in Clinical Epidemiology
Likelihood of Antimicrobial Resistance in Urinary E. coli Isolates Comparing Female Patients with Recurrent Versus Non-recurrent Uncomplicated Urinary Tract Infection in the United States
Background: The prevalence of urinary tract infection (UTI) caused by extended spectrum β-lactamase-producing Enterobacterales, as well as antimicrobial resistant (AMR) and multi-drug resistant uropathogens, is increasing in female patients with uncomplicated UTI (uUTI). However, there is a paucity of data concerning AMR in recurrent uUTI (RuUTI) infections.
Objectives: The study objective was to assess the likelihood of AMR in patients with RuUTI versus those with non-recurrent uUTI (NRuUTI).
Methods: Data from female patients ≥12 years of age were assessed in the OPTUM de-identified Electronic Health Record (EHR) dataset between October 2015 and February 2020. Eligibility criteria included ≥1 culture-proven E. coli uUTI diagnosis and pyuria ±7 days of diagnosis (date of most recent uUTI diagnosis was defined as index date), treatment with ≥1 oral antibiotic of interest ±5 days of index date, susceptibility test results of the urinary E. coli isolate to ≥3 antibiotic classes of interest ±7 days of index date, and ≥12 months of EHR activity prior to index (baseline period). Patients were categorized into 2 cohorts: RuUTI (defined as ≥2 uUTI episodes in 6 months or ≥3 uUTI episodes in 12 months including index diagnosis) and NRuUTI. Multivariable logistic regression was performed to evaluate the relative odds of AMR in RuUTI versus NRuUTI.
Results: Of 80,267 patients with uUTI, 12,234 (15.2%) were included in the RuUTI cohort and 68,033 (84.8%) in the NRuUTI cohort. Mean (standard deviation) age was 51.2 (21.0) and 46.5 (19.5) years (standardized difference [std diff] 13.1%), and 45.5% and 34.6% had a history of acute or semi-acute infections (non-UTI related) at baseline (std diff 22.3%), respectively. Patients with RuUTI had 18% (95% CI 1.12, 1.24) higher odds of single drug resistance (SDR; p< 0.001), 53% (95% CI 1.41, 1.67) higher odds of resistance to 2 drug classes (MDR2; p< 0.001), and 70% (95% CI 1.48, 1.96) higher odds of resistance to ≥3 drug classes (MDR3) relative to patients with NRuUTI (p < 0.001).
Conclusions: Patients with RuUTI due to E. coli had higher odds of AMR than patients with NRuUTI. The magnitude of the odds ratio increased from SDR to MDR3. The increased likelihood of AMR in RuUTI should be considered when treating patients empirically with antibiotics.